Objective To explore the diagnostic performance of a spectral CT-based radiomics machine learning model and nomogram for preoperatively identifying the KRAS gene status in patients with colorectal cancer (CRC). Methods A total of 137 CRC patients who underwent KRAS mutation detection and preoperative spectral CT examination were retrospectively included (70 cases with KRAS wild type and 67 cases with KRAS mutant type). They were randomly divided into a training set (95 cases) and a test set (42 cases) in a 7∶3 ratio. Tumor region of interest (ROI) was delineated on venous-phase 70 keV monochromatic enhanced CT images, and radiomics features were extracted and selected. A radiomics score (Rad-score) was calculated using least absolute shrinkage and selection operator (LASSO) regression. Six models were established including three radiomics models based on support vector machine (SVM), extreme gradient boosting (XGBoost), and logistic regression (LR), as well as three combined models integrating spectral CT imaging features with the Rad-score. Model performance was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC), and compared using the Delong test. A radiomics nomogram was constructed based on the Rad-score and validated in the test set. Calibration curves, decision curve analysis (DCA), and clinical impact curves were used to assess calibration, clinical net benefit, and clinical utility. Results A total of 8 radiomics features and 1 spectral parameter were selected. In the test set, the LR-based combined model demonstrated the best performance, with an AUC of 0.891, outperforming the combined models based on SVM (AUC=0.796), XGBoost (AUC=0.787), and LR (AUC=0.812) (all P<0.05), as well as the combined models based on SVM (AUC=0.889) and XGBoost (AUC=0.873) (both P<0.05). The nomogram model achieved AUCs of 0.987 and 0.916 in the training and test sets, respectively. The calibration curve showed good agreement in the training set, while performance in the test set was slightly lower. DCA and clinical impact curves demonstrated that the nomogram provided favorable clinical net benefit and utility. Conclusion The LR-based model and nomogram, constructed using venous-phase spectral CT and radiomics features, offer valuable preoperative insights into KRAS gene status in CRC patients and may serve as a reference for clinical decision-making.

Objective To investigate the association between the pericoronary adipose tissue fat attenuation index (FAI) surrounding culprit plaques in acute coronary syndrome (ACS) and plaque characteristics, and to assess its value in predicting culprit plaques. Methods This retrospective study enrolled 50 patients diagnosed with ACS (ACS group) and 40 asymptomatic individuals with coronary atherosclerosis who underwent coronary computed tomography angiography (CCTA) during the same period (control group). Clinical and imaging data were analyzed. In the ACS group, plaques were classified as culprit or non-culprit plaques. Based on the number of high-risk features, plaques were further categorized as non-high-risk or high-risk. FAI surrounding plaques was measured using predefined default (-190 to -30 HU) and wide (-190 to 20 HU) attenuation thresholds. Student’s t-test, one-way ANOVA, and chi-square test were used to compare FAI values of plaques with different characteristics and degrees of stenosis between and within groups; the plaque characteristics, stenosis severity, and FAI among culprit plaques, non-culprit plaques, and control group plaques; the high-risk features between culprit and non-culprit plaques; and the FAI values between high-risk and non-high-risk plaques. Multivariable logistic regression analysis was performed to identify independent predictors of culprit plaques. Receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive performance of individual and combined factors for culprit plaques. The DeLong test was used to compare the differences in the area under the curve (AUC) among individual and combined factors. Results The FAI measured with the wide threshold was significantly higher than that measured with the default threshold for culprit plaques, non-culprit plaques, and control group plaques (all P<0.05). Under both thresholds, the FAI of culprit plaques was significantly greater than that of non-culprit plaques and control plaques (all P<0.05). Among the culprit plaques, 64% were classified as high-risk plaques, and these also showed high proportions of mixed plaque morphology, severe stenosis, and occlusion (52%, 76%, and 12%, respectively). In the ACS group, the FAI surrounding calcified plaques was lower than that surrounding non-calcified and mixed plaques (P<0.05). The FAI was significantly higher around plaques causing severe stenosis or occlusion (P<0.05), and higher around high-risk plaques compared to non-high-risk plaques (P<0.05). Multivariable logistic regression analysis indicated that stenosis severity ≥ moderate, higher default threshold FAI, and a greater number of high-risk plaque features were independent predictors of culprit plaques. The combination of default threshold FAI, stenosis severity, and high-risk features yielded the highest predictive performance (AUC=0.981). DeLong test analysis showed that the AUCs of models combining default threshold FAI with other factors were significantly higher than those of any single factor alone (all P<0.05). Conclusion The FAI surrounding ACS plaques can partially reflect plaque inflammation and vulnerability. Combining default threshold FAI with stenosis severity and high-risk features improves diagnostic performance in identifying culprit plaques.

Preoperative prediction of the efficacy of neoadjuvant immunotherapy (NIT) in non-small cell lung cancer (NSCLC) helps identify patients who are likely to benefit, reduce the risk of postoperative recurrence and metastasis, and improve prognosis. Radiomics and deep learning can be used to explore imaging biomarkers for predicting NIT efficacy in NSCLC. Radiomics, through global feature analysis or habitat analysis methods, can effectively quantify the temporal and spatial heterogeneity of tumors, providing a quantitative basis for efficacy prediction. Deep learning, on the other hand, adaptively extracts deep imaging features to evaluate treatment response. This review summarizes recent research progress in radiomics and deep learning technologies for predicting NIT efficacy in NSCLC patients, and discusses the associated technical challenges and corresponding solutions.

Ultrafast dynamic contrast-enhanced (UF-DCE) MRI, with its advantages of high imaging speed, high temporal resolution, and the ability to obtain rich hemodynamic parameters, has been utilized in the early screening, differential diagnosis, neoadjuvant chemotherapy efficacy prediction, and prognostic evaluation of breast cancer. This review summarizes the technical principles of UF-DCE MRI, its applications in breast cancer diagnosis and treatment, and the research progress on artificial intelligence applications in UF-DCE MRI.

Amyloid-related imaging abnormalities (ARIA) are among the most common adverse reactions during Aβ monoclonal antibody treatment for early Alzheimer’s disease (AD). Magnetic resonance imaging (MRI) serves as a crucial tool tool for monitoring the occurrence of ARIA and assessing its severity throughout the treatment process. This paper provides a detailed overview of the mechanisms of ARIA, its imaging manifestations, and severity grading. Additionally, a comprehensive standard MRI examination protocol and monitoring workflow are proposed based on clinical practice experience. The study also highlights the critical role of imaging monitoring in guiding clinical medication for AD patients.

Body fat percentage serves as a crucial indicator for measuring an individual’s body fat content. Ultrasound, as a safe and non-invasive examination method, not only enables visual detection of fat layer thickness and effective differentiation between subcutaneous and visceral fat, but also allows assessment of body fat percentage through quantitative measurements of fat thickness at multiple sites. Notably, appropriate selection of measurement sites is particularly beneficial for improving the accuracy of body fat percentage estimation. This review summarizes the potential value, reproducibility, and application ultrasound in measuring subcutaneous and visceral fat for body fat percentage assessment.

Habitat imaging(HI) can analyze tumor heterogeneity and microenvironmental characteristics and has been increasingly applied in the research, diagnosis, and treatment of common digestive system tumors, including colorectal cancer, gastric cancer, and hepatocellular carcinoma. Currently, HI is used to construct genotypic prediction models, precision staging, and metastasis prediction in colorectal cancer; to quantify immune microenvironment characteristics, evaluate treatment response, and predict prognosis in gastric cancer; and to achieve non-invasive identification of microvascular invasion and recurrence risk stratification in hepatocellular carcinoma. This article introduces the basic principles and technical processes of HI, and reviews its research progress in the above-mentioned digestive system tumors.

Early diagnosis and accurate assessment of kidney damage are crucial for the treatment and prognosis of chronic kidney disease (CKD). Radiomics can deeply mine information from medical images and extract a large number of quantitative features that are not recognizable by the human eye, thereby constructing models for the diagnosis and staging of CKD, as well as evaluating kidney function and the degree of renal fibrosis. This paper reviews the research progress of radiomics based on ultrasound, MRI, and CT in the diagnosis and evaluation of CKD.

In 2024, the American Heart Association (AHA) issued a scientific statement on the diagnosis and management of cardiac sarcoidosis (CS), systematically outlining the diagnostic and therapeutic framework for this infiltrative cardiomyopathy characterized by non-necrotizing granulomatous inflammation. The Statement emphasizes that multimodal imaging is a core pillar of CS diagnosis. This article focuses on the core content of the statement, specifically interpreting the clinical application value and diagnostic standards of multimodal imaging techniques for CS, as well as the collaborative diagnostic and therapeutic strategies involving cardiac magnetic resonance (CMR) and positron emission tomography (PET), with the aim of providing precise imaging assessment pathways for clinical practice.

Objective To explore the value of a machine learning model incorporating primary tumor and peritumoral radiomics features for preoperative prediction of lymphovascular invasion (LVI) in gastric cancer. Methods Clinical and imaging data of 148 patients with pathologically confirmed gastric cancer were retrospectively collected. Based on pathological results, patients were divided into an LVI-positive group (79 cases) and an LVI-negative group (69 cases). Patients were randomly divided into a training set (103 cases) and a test set (45 cases) in a 7∶3 ratio. Radiomic features were extracted from the primary tumor and peritumoral regions. The least absolute shrinkage and selection operator (LASSO) method was used to select optimal radiomic features, and the radiomics score (Rad-score) was calculated. The clinical features with statistically significant differences between the two groups were combined with Rad-score for multivariate logistic regression analysis to select variables for constructing a machine learning model. Seven machine learning algorithms, including logistic regression (LR), extreme gradient boosting (XGBoost), random forest (RF), Gaussian naive Bayes (GNB), support vector machine (SVM), light gradient boosting machine (LightGBM), and K-nearest neighbors (KNN), were used to construct clinical-radiomics models. The performance of the models was evaluated using receiver operating characteristic (ROC) curve analysis. Calibration curves and decision curve analysis (DCA) were used to assess the calibration degree and clinical net benefit of the models, respectively. The SHapley Additive exPlanations (SHAP) method was employed to provide visual interpretation of the predictive model. Results In the training set, all seven machine learning models achieved an AUC greater than 0.650, with the RF model achieving the highest AUC (0.858), sensitivity (0.895), and accuracy (0.776). The calibration curve indicated that the RF model had the lowest Brier score (0.153), demonstrating the best predictive accuracy. DCA revealed that the RF model provided the highest net clinical benefit when the risk threshold ranged from 0.30 to 0.70. In the test set, the RF model maintained stable diagnostic performance, achieving an AUC of 0.821. SHAP analysis identified key factors associated with LVI risk in gastric cancer patients and provided visual interpretation for individual predictions. Conclusion The RF model, integrating primary tumor and peritumoral radiomic features with clinical factors, holds significant value for preoperative prediction of LVI status in gastric cancer patients.

Chronic obstructive pulmonary disease (COPD) is a common and preventable condition, and early diagnosis and intervention are crucial for improving patient outcomes. Pulmonary CT imaging enables quantitative analysis of lung density, airways, vessels, and body composition, revealing COPD phenotypes and identifying comorbidities. This approach aids in assessing the severity of COPD and facilitating early intervention. This review summarizes the research progress on quantitative chest CT analysis in COPD.

Objective To explore the feasibility of radiomics models based on contrast-enhanced computed tomography (CECT) in distinguishing cephalic para-carcinoma tissues and resection margin in esophageal squamous cell carcinoma (ESCC) following neoadjuvant chemotherapy and immunotherapy (NACI). Methods This retrospective study included 188 pathologically confirmed ESCC patients who underwent NACI, recruited from two medical centers. A total of 138 patients from Center A were randomly divided into a training set (97 cases) and an internal validation set (41 cases) at a 7∶3 ratio, while 50 patients from Center B served as an external validation set. Using an open-source software 3D-Slicer, four regions of interest (ROIs) representing cephalic para-carcinoma tissues (P1, P2, P3, and P4) at 1 cm, 2 cm, 3 cm, and 4 cm above the tumor margin, respectively, and one ROI for resection margin tissue (P5, 5 cm above the tumor) were delineated on CECT images. Radiomics features were extracted using the Pyradiomics package. The radiomics features obtained from four cephalic para-carcinoma tissues were individually paired with those of resection margin tissue to differentiate between them, which were designated as groups P1, P2, P3, and P4, respectively. Univariate analysis and the least absolute shrinkage and selection operator (LASSO) method were employed to select optimal radiomics features in the training sets, and logistic regression models were constructed. The area under the receiver operating characteristic (ROC) curve (AUC) was used to assess the discriminatory performance of the radiomics models. Results The AUCs of the P1 model in the training, internal validation, and external validation sets were 0.831, 0.820, and 0.787, respectively. The AUCs of the P2 model were 0.809, 0.797, and 0.769, respectively. Both the P1 and P2 models demonstrated good discriminatory performance (AUC>0.76), with the P1 model achieving higher AUC values than the P2 model in all datasets. Conclusion The CECT-based radiomics model demonstrates high efficacy in distinguishing cephalad peritumoral (P1 and P2) and resection margin tissues in ESCC following NACI.

Objective To explore the predictive value of a radiomics-based nomogram using ¹⁸F-FDG PET for KRAS gene mutation in colorectal cancer (CRC). Methods A total of 115 patients with CRC who underwent ¹⁸F-FDG PET/CT before treatment were enrolled retrospectively, with an average age of 56.40±1.17 years. According to the gene phenotype, the patients were divided into a KRAS mutant group (56 cases) and a KRAS wild-type group (59 cases). All patients were randomly assigned into a training set of 80 cases and a validation set of 35 cases at a 7∶3 ratio. Lesion regions of interest (ROI) were delineated on PET images using ITK-SNAP, and radiomics features were extracted using the Pyradiomics package. Feature selection was performed using the least absolute shrinkage and selection operator (LASSO), and radiomics scores (Rad-scores) were calculated to construct the radiomics model. Univariate analysis was conducted on clinical data, and statistically significant variables were combined with Rad-scores for multivariate logistic regression to identify independent predictors of KRAS mutations. Clinical models and combined models were developed, with the latter presented as a nomogram. The predictive performance of the models was evaluated using receiver operating characteristic (ROC) curves, and the differences in performance were compared using the Delong test. Decision curve analysis (DCA) was used to assess the clinical utility of the nomogram, and a calibration curve was employed to evaluate its calibration accuracy. Results In both the training and validation sets, the combined model demonstrated the highest predictive performance, with area under the curve (AUC) values of 0.787 and 0.727, respectively, which were significantly higher than those of the clinical model (Z=-2.230, P=0.026; Z=-2.223, P=0.026). DCA showed that the nomogram provided clinical net benefits when the threshold probability ranged from 0.13 to 0.92 in the training set and from 0.08 to 0.93 in the validation set. The calibration curve indicated good agreement between predicted and actual values for the nomogram. Conclusions The radiomics-based nomogram using ¹⁸F-FDG PET imaging holds potential predictive value for KRAS gene mutation in CRC, and may assist clinicians in devising individualized treatment strategies.

Objective To compare the dosimetry and efficacy differences among Hyperarc, volume modulated arc therapy (VMAT), intensity modulated radiation therapy (IMRT), and three-dimensional conformal radiation therapy (3DCRT) for the treatment of multiple brain metastases. Methods A total of 168 patients with multiple brain metastases were prospectively selected and randomly divided into 4 groups (n=42), each receiving treatment plans designed using Hyperarc, VMAT, IMRT, or 3DCRT. One-way ANOVA, Kruskal Wallis test, and chi-square test were used to compare the clinical data and treatment efficacy, as well as the dosimetric parameters of the planning target volume [homogeneity index (HI), conformity index (CI), gradient index (GI)], organ at risk dose [maximum dose (Dmax) and average dose (Dmean) to bilateral crystals and brainstem], and monitor units and beam delivery time, among the 4 groups. Results The treatment efficacy of the Hyperarc group was significantly higher than that of the VMAT, IMRT, and 3DCRT groups (P<0.05). Compared with the other three groups, the Hyperarc group had lower HI, GI, and Dmax/Dmean for the bilateral lenses and brainstem, while Dmean and CI of target volume were higher (all P<0.05). The VMAT and IMRT groups exhibited lower HI, GI, and Dmax/Dmean for the bilateral lenses and brainstem compared to the 3DCRT group (all P<0.05), while their target volume Dmean and CI were higher (P<0.05). The Hyperarc group had fewer monitor units and a shorter beam-on time than the VMAT and IMRT groups, but higher than the 3DCRT group (all P<0.05). Conclusion In the radiotherapy of multiple brain metastases, Hyperarc technology demonstrates advantages in target volume dosimetric distribution and organ-at-risk protection.

Coronary computed tomography angiography (CCTA) is the preferred non-invasive imaging modality for the diagnosis, clinical treatment decision-making, therapeutic efficacy assessment, and prognosis prediction of coronary artery disease (CAD). CCTA can be used not only to evaluate plaque characteristics and the degree of luminal stenosis but also to provide quantitative parameters reflecting coronary flow reserve, peri-coronary fat inflammation, left ventricular myocardial strain, myocardial fibrosis, and myocardial perfusion. This review summarizes the application of these quantitative derived parameters in CAD.

Early screening and monitoring of osteoporosis (OP) are key to preventing fragility fractures. Various MRI quantitative techniques, including water-fat separation imaging (IDEAL-IQ, 3D mDixon Quant), vertebral bone quality (VBQ) scoring, ultrashort echo time (UTE), high-resolution MRI (HR-MRI), intravoxel incoherent motion (IVIM), dynamic contrast-enhanced MRI (DCE-MRI), and magnetic resonance ensemble sequences, have demonstrated high efficacy in quantitatively evaluating bone marrow fat-water composition, trabecular microstructure, bone marrow microcirculatory perfusion, and biophysical properties in OP patients. This review summarizes the research progress of the these MRI quantitative techniques in OP assessment.

Objective To investigate differences in white matter microstructure between patients with neuropsy-chiatric systemic lupus erythematosus (NPSLE) and those with non-neuropsychiatric SLE (Non-NPSLE) using tract-based spatial statistics (TBSS). Methods A total of 34 NPSLE patients and 32 Non-NPSLE patients were prospectively enrolled, along with 33 healthy controls (HC) during the same period. All participants underwent brain diffusion tensor imaging (DTI). TBSS was used to compare white matter microstructural differences among the three groups. Fractional anisotropy (FA) values were compared using one-way ANOVA, with post hoc analyses conducted for pairwise group comparisons. Partial correlation analyses assessed the relationships between FA values of significantly different clusters and neuropsychological scores or clinical indicators, as well as the correlations between neuropsychological scores and clinical indicators. Results Five clusters showed significant FA differences among the three groups (P<0.05, FWE-corrected). Post hoc analysis revealed that two clusters in both the Non-NPSLE and NPSLE groups had lower FA values than the HC group, and one cluster in the NPSLE group had a lower FA value than the Non-NPSLE group (P<0.05, FWE-corrected), indicating more extensive white matter involvement in NPSLE. FA reductions in SLE patients were primarily located in the corpus callosum and corona radiata. Correlation analysis showed that FA values of the significant clusters in pairwise comparisons were positively correlated with IgM levels (P<0.05). In the NPSLE group, HADS-D scores were negatively correlated with C4 levels (r=-0.354, P=0.047), while in the Non-NPSLE group, MoCA scores were negatively correlated with ESR (r=-0.424, P=0.019). Conclusion NPSLE patients exhibit more extensive white matter microstructural damage compared to Non-NPSLE patients. The FA values of some differential clusters correlate with clinical indicators, suggesting that these changes may serve as important imaging biomarkers for detecting disease activity or neuropsychiatric involvement in SLE.

Objective To explore the feasibility of training a deep learning model for fully automated adrenal segmentation on non-contrast CT images. Methods The images and clinical data of 1 200 patients who underwent non-contrast adrenal CT scan were retrospectively collected. Using a 5-fold cross-validation method, patients were divided into a training set (960 cases) and an internal test set (240 cases) at an 8∶2 ratio. Additionally, 81 cases who underwent adrenal CT scans were collected as an independent test set. Both 2D nnU-Net and 3D nnU-Net segmentation models were constructed based on the nnU-Net framework. Clinical and CT imaging features were compared between the two groups using the Mann-Whitney U test and chi-square test. The model’s segmentation performance was objectively evaluated using the Dice coefficient (DSC), Hausdorff distance (HD), average symmetric surface distance (ASSD), recall, and precision from the internal testing set and independent testing set. Two radiologists subjectively evaluated the 3D nnU-Net segmentation results on the independent test set. Results No statistically significant differences were observed in general characteristics between the training set+internal test set and independent test set (all P>0.05). Both 2D and 3D nnU-Net models achieved high segmentation performance for the left and right adrenal glands on the internal and independent test sets. Compared to the 2D nnU-Net model, the 3D nnU-Net model demonstrated higher DSC and precision, lower HD and ASSD, and similar or higher recall. The segmentation results of the 3D nnU-Net were closer to manual annotations compared to the 2D nnU-Net model. Subjective evaluation by two radiologists on the independent test set revealed 62.96% satisfactory and 37.04% unsatisfactory segmentation outcomes for the 3D nnU-Net. Conclusion The deep learning-based adrenal segmentation mode is feasible for automatic adrenal segmentation on non-contrast CT images.

Heart failure with preserved ejection fraction (HFpEF) is a condition characterized by high heterogeneity and complex etiologies. Among its various underlying mechanisms, myocardial fibrosis is a key pathophysiological factor contributing to cardiac dysfunction in HFpEF patients. This article reviews the progress in applying various cardiac magnetic resonance (CMR) techniques, including delayed gadolinium enhancement (LGE), T₁ mapping, extracellular volume (ECV), and CMR feature tracking (CMR-FT), to assess myocardial fibrosis in HFpEF. It highlights the role of these techniques in evaluating the distribution and severity of myocardial fibrosis, as well as their utility in risk stratification and prognostic assessment. Additionally, the article explores the future potential of artificial intelligence in enhancing CMR-based evaluations.

Neoadjuvant chemotherapy (NAC) can not only reduce the stage of breast cancer but also enable some tumor lesions and axillary lymph nodes to achieve pathological complete response (pCR). Accurate preoperative imaging assessment of axillary lymph node status after NAC in breast cancer patients can help avoid excessive surgical intervention and guide the development of individualized treatment plans. This review summarizes recent progress in the use of imaging methods such as ultrasound, MRI, CT, and PET/CT to evaluate axillary lymph node pCR after NAC.

Vascular lesions of the head and neck are one of the major contributing factors to serious health issues such as stroke, with pathological progression closely linked to hemodynamic abnormalities. Computational fluid dynamics (CFD), through patient-specific three-dimensional vascular modeling, enables the quantitative analysis of geometric parameters and hemodynamic characteristics. This approach elucidates the pathophysiological mechanisms of carotid atherosclerosis and intracranial aneurysms, playing a crucial role in predicting disease progression and guiding therapeutic strategies. This review systematically summarizes recent advancements in CFD applications for cerebrovascular pathologies of the head and neck region.

Imaging assessment plays a crucial role in the follow-up of patients after radical surgery for colorectal cancer. Fibroblast activation protein inhibitor (FAPI), as an emerging radionuclide tracer, is not affected by glucose metabolism and offers a higher tumor-to-background ratio, making it more effective in detecting recurrent lesions. It is particularly advantageous in the detection of non-FDG-avid tumors, peritoneal metastases, and small lesions. This review summarizes the advances in the application of FAPI PET/CT for assessing recurrence and metastasis after radical surgery for colorectal cancer.

Vulnerable carotid plaques and cerebral small vessel disease (CSVD) are both significant etiologies of ischemic stroke, and they are closely interrelated. Imaging modalities, including ultrasound, CT angiography (CTA), and MR high-resolution vessel wall imaging (HR-VWI), enable the assessment of various vulnerable plaque characteristics. This review focuses on the correlation between vulnerable plaques and CSVD, aiming to provide a theoretical basis for further research and clinical application.

Objective To investigate the relationship between pericoronary fat attenuation index (FAI) and the plaque progression. Methods This retrospective study included 140 inpatients with suspected coronary artery disease (CAD) who underwent consecutive coronary computed tomography angiography (CCTA), with an average age of 56.8 ± 10.5 years old. A total of 348 plaques were identified. Clinical data, as well as baseline and follow-up (at least one year apart) CCTA imaging data, were collected. Characteristics, degree of stenosis, plaque volume (PV), and percentage atheroma volume (PAV) were analyzed. Peri-plaque FAI was measured and analyzed. A multivariate generalized estimating equation was used to adjust for confounding variables, and linear regression models were fitted to analyze the association between the annual change in FAI (ΔFAI/y) and the annual changes in PV (ΔPV/y) and PAV (ΔPAV/y). Results The median interval between the two CCTA scans was 2.3 (1.7, 3.8) years. Quantitative analysis of the two scans revealed significant increases in all components of PV and PAV except for lipid PV and lipid PAV (all P<0.001). After adjusting for China-PAR score, Leiden score, TyG index, antiplatelet therapy, and statin use using generalized estimating equation, ΔPAV/y of total plaques was significantly positively correlated with ΔFAI/y (β=0.156, 95%CI: 0.025-0.287, P=0.019). Specifically, ΔPAV/y of non-calcified plaques (β=0.139, 95%CI: 0.006-0.273, P=0.041) and fibrous plaques (β=0.197, 95%CI: 0.067-0.327, P=0.003) also showed significant positive correlations with ΔFAI/y. Conclusion Changes in FAI are consistent with changes in non-calcified PAV and fibrous PAV. This may help identify potentially high-risk patients with stable coronary artery disease and supports the use of FAI as a valuable tool for evaluating treatment efficacy in coronary artery disease.

Multimodal breast MRI is a vital non-invasive tool for evaluating breast cancer. MRI-based radiomics can extract high-throughput and in-depth features of breast cancer tissues, uncovering relationships with the expression levels of breast cancer molecular markers. This review summarizes the current research status of artificial intelligence in predicting breast cancer molecular subtypes from multiple perspectives, including the basics of MRI modalities, advances in radiomics research, machine learning and deep learning algorithms and their clinical applications, and multi-omics combined studies. It highlights the advantages and limitations of different methods, providing a reference for future MRI-based artificial intelligence research and the non-invasive clinical prediction of breast cancer molecular subtypes.

Patients with advanced non-small cell lung cancer (NSCLC) exhibit variable responses to immunotherapy. Therefore, the early and accurate prediction of immunotherapy efficacy is crucial. Currently, artificial intelligence (AI) based on CT imaging are widely utilized for predicting immunotherapy outcomes in advanced NSCLC. This review summarizes the research progress and challenges of AI applications based on CT imaging, specifically in the context of immunotherapy for advanced NSCLC, including the prediction of therapeutic efficacy and prognosis as well as the detection of adverse effects. It also analyzes future research directions and potential development prospects.

The diagnosis and treatment of refractory epilepsy are challenging and often associated with generally poor prognoses. Integrated PET/MRI combines the advantages of PET and MRI imaging modalities, enabling the simultaneous acquisition of structural, functional, and molecular-level information. This comprehensive and objective approach provides valuable diagnostic and therapeutic insights for patients with refractory epilepsy. This article briefly outlines the imaging principles of integrated PET/MRI，and reviews its clinical applications in the preoperative localization and postoperative follow-up evaluation of refractory epilepsy.

Ischemia with non-obstructive coronary artery disease (INOCA) is characterized by coronary microvascular dysfunction and epicardial vasospasm as its core pathological mechanisms, which significantly increase the risk of adverse cardiovascular events. Non-invasive imaging modalities, including cardiac magnetic resonance imaging, myocardial computed tomography perfusion imaging, echocardiography, and positron emission tomography, have demonstrated significant potential in the evaluation of INOCA. Artificial intelligence further enhances the efficiency and accuracy of imaging-based assessments. This article provides a systematic review of the pathophysiological mechanisms underlying INOCA and recent advances in imaging techniques for its evaluation, with a focus on clinical applicability and technological innovation.